Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?

To: Bill Jackson who wrote (73872)	3/7/2002 12:16:58 PM
From: semiconeng	Read Replies (1) \| Respond to of 275872

Semi, Do they use the same mask for a small area and then used it again and again..... -- Same mask, stepped to the next location for the same layer. Different Mask for different layers. Logically a smaller sized CPU would get more on the mask. Let us say this mask holds a 3x3 grid of CPUs at .18 and a 4x4 grid at .13(these are arbitrary picks). Since it would take the same number of step and repeats to do the wafer unless exposure times varied with feature size?) then the throughput would be the same. -- I think I agree with this.... Partially. Let's use the assumption of the 3x3 and 4x4. Usually, if you can get an extra die on a mask, it would be better to orient them vertically, so that you would shoot an additional "row" of die with each full pass of the wafer. Step/Scan shoots a row (or 2, or 3) left to right, steps down and shoots right to left (with wafer Notch oriented down). Shrinking the geomitries does allow more die on the mask (sometimes), with more die on the wafer (always), so it looks like a wash, but again, shorter scanning times of the smaller individual die should result in increased throughput, with number of steps staying the same, or even increasing. But I could be wrong, like I said, I have no throughput data, just a WEG. It is also possible that with smaller features the machine takes more time to move and align for the exposure due to the smaller features? --Naaaahhh. For the Step/Scan tools that I've seen, Alignment is done at the beginning of the wafer shot, when the wafer is placed on the chuck. It works something like: Install specific reticle, Align Reticle to tool, Place wafer, Align Wafer to Reticle, Shoot. Since the Wafer and Reticle are moving in sync during the shot, and the exact die size is known, there's no need to realign the Reticle/Wafer die to die. Just move the Reticle/Wafer to the exact spot of the next die with stepper motors. The only thing that I know of that happens during the shot from die to die, is Focus Adjustment. Automatic Dynamic Focusing on the fly.... Man, now THAT was advanced technology.... If they made one huge mask that covered the whole wafer then edge effects would make problems and you may have to optimize the focus to minimize edge loss and that give you center losses? -- If you did that, I THINK that you would also need a bigger lens to cover the whole wafer. Man, I don't even want to think about the optical problems that would exist trying to make a lens that big. Lens resolution uniformity is one of the bigger things holding back smaller profiles, but yes, I would guess there would be a trade off from center of lens to edge. Is it any better with reflective lenses versus transmissive lenses? bill --Ya got me.... I'm no longer a "Litho Puke" (Our fond Etch name for them), so my general lens knowledge is probably still applicable, but material lens specifics are beyond my current knowledge. Do ya want me to ask the "pukes"? :-) Semi